چكيده به لاتين
Abstract:
One of the most important data mining techniques is scoring method, to set profit and guarantee rate consider index and standards and customers who have suitable credit position can benefit from that.
In order to consider the right and Islamic justice, it’s worthy, valued that credit brokers are diligent be more diligent in order to more use and decreasing probable dangers.
Today’s according to the volume of deals from people who receive facilities and financial crisis, reduplicate identification and banks risk management. As mentioned, in this research we’re trying to survey 350 banks customers and with offered method can find suitable pattern for customer risk prediction.
Risk in bank operation can threaten the power of the bank profits. This threat can destroy part of bank deposits and even part of bank fund.
In this research, we’re trying to present one model of costumers’ validation for award appropriate facility to each class based on genetics algorithm. Genetics algorithm with proper features selection can help costumers’ validation to make optimized decision tree for decomposition data analysis that neural network method and genetic algorithm can use them for tutorial. The main purpose in this issue is with use of decision tree can optimally categorize loyal and unloyal bank’s customers.
Keywords: Data Mining, Scoring, Feature Selection Algorithm, Genetics Algorithm, Validation, Bank
